Hello guys, this is not really a VDA topic but more of a general topic. We are using persistent single-session VDAs as "virtual personal Desktops". There are about 250 of those machines in our environment and all of them are assigned to individual users. We often have the problem that they have low disk space and need to cleanup regularily and sometimes can't even do Windows-Updates. In regards of the current situation ordering more storage space is very expensive and can take up some time - so we are currently thinking about implementing an automatic cleanup for these clients over ControlUp or SCCM Baselines. Did any of you guys do something similar in the past? What are good files/paths which can be cleaned up regularily? Thanks for your answers

Depends on what is causing your low disk space. A persistent machine with a local profile WILL generate lots of crud. I'd suggest using some sort of pruning script to reduce caches at logoff. There are a few kicking about on the interwebs

I would use Treesize tool to quickly analyze root cause of folder sizes. Optimization tweaks might interfere with purging previous installed Windows Updates. Perhaps use desktop cleanup wizard to double check. Once you know the apps that cause large cache data you can check what folders can be purged. But if you already know root cause, then you should use logoff cleanup script as James mentioned.

100% agree with the gentlemen above.. persistent with no cleanup is like your partner adding more clothes to the cabinet without ever throwing something old out.. at some point it causes storage issues.

On persistent personal VDAs the bloat is rarely the profile, it is system crud and the WU pipeline, which is also why patching starts failing. Profile a sample with TreeSize first, then target these high-yield paths: • Component store: DISM /Online /Cleanup-Image /StartComponentCleanup (avoid /ResetBase unless you accept losing update rollback) • C:\Windows\SoftwareDistribution\Download (stop wuauserv, clear, restart) and the Delivery Optimization cache • CCMcache (prune via the ConfigMgr resource COM API, never raw delete) • C:\Windows\Temp, %LOCALAPPDATA%\Temp, Teams (classic + new), Edge/Chrome and OneDrive caches • CBS logs, WER, thumbnail/icon cache, old shadow copies if enabled For automation: ship it as an SCCM Configuration Baseline (detection + remediation script) so you get compliance reporting and self-healing across all 250, rather than a one-shot logoff script. Keep ControlUp for a free-space alert around 15% so WU never gets blocked again. Storage Sense is a fine backstop but not the primary mechanism, and Geert is right that aggressive WU purging can break servicing, so test StartComponentCleanup cadence before rolling wide. Clear the component store and WU staging first, that reclaims more than anything else and unblocks your updates.

Hard disagree, but I work with developers and bankers